Inside Apple’s secret plan to build a cloud platform and take on AWS, Google Cloud

Apple has reportedly considered launching its own developer-focused cloud service, powered by its custom silicon, which could position it in direct competition with Amazon Web Services (AWS), Microsoft Azure, and Google Cloud. According to a detailed investigation by The Information’s Aaron Tilley, the initiative, which was internally referred to as Project ACDC (Apple Chips in Data Centres), was discussed actively within Apple through the first half of 2024.

But what would such a move mean for Apple and the broader developer ecosystem? And why now?

The Cloud Market: Massive, Profitable, and Competitive

Cloud infrastructure services are a foundational part of the internet today. AWS alone generated $25 billion in revenue in Q1 2024, according to Amazon’s earnings report. Microsoft Azure and Google Cloud are also growing rapidly, with Microsoft’s Intelligent Cloud segment reaching $26.7 billion in the same quarter.

These platforms allow businesses and developers to host applications, store data, and run compute-intensive tasks such as AI training and inference without managing physical servers.

What Is Apple’s Project ACDC?

As reported by MacRumors in 2024, Apple internally explored the idea of letting developers rent access to servers powered by its M-series chips, similar to those used in Macs and iPads. These chips are built on the Arm architecture and are known for their power efficiency and AI inference capabilities.

Apple already uses its own silicon in some of its data centres, particularly to power Private Cloud Compute, the secure server-side backbone for its newly announced Apple Intelligence features. This infrastructure processes user data in a privacy-preserving way and leverages Apple’s Neural Engine for inference at scale.

The same report notes that Apple initially tested these Mac chip-powered servers with the Siri team for text-to-speech capabilities, before expanding their use across services like Apple Music, Photos, and Wallet.

Why Would Apple Consider Entering This Market?

1. Services Revenue Is Under Pressure

Apple’s Services business generated $85.2 billion in FY2023, according to the company’s annual report. But with increasing regulatory scrutiny over its App Store commissions in the EU and the United States and ongoing antitrust investigations into its $20 billion per year Google search deal, Apple is under pressure to diversify its revenue streams. A developer-facing cloud platform could provide a fresh avenue for growth.

2. Maximising Its Silicon Advantage

Apple’s chip division has been one of its biggest technological differentiators. If Apple can offer comparable or better performance for cloud-based AI inference at a lower energy and hardware cost, that becomes a compelling value proposition, especially as AI adoption grows.

3. Controlling the Stack

Apple’s long-term strategy has consistently been about controlling hardware, software, and services. A cloud service powered by Apple Silicon would give the company more end-to-end control over performance, privacy, and developer experience.

What Might Apple’s Cloud Look Like?

There is no official announcement, but based on current reports, a developer-centric Apple Cloud could:

• Integrate tightly with tools like Xcode, Swift, and Core ML
• Allow seamless deployment of on-device machine learning models at scale
• Provide a more private, Apple-like alternative to third-party platforms like AWS
• Possibly fall under the iCloud umbrella, though with a dedicated developer tier

Importantly, Apple reportedly did not plan to create a traditional enterprise sales division. Instead, its Developer Relations team was considered for managing access, creating a simpler onboarding experience compared to AWS or Azure (1).

Technical Focus: AI Inference, Not Training

Apple’s M-series chips and Neural Engine are designed for on-device inference, which involves running pre-trained AI models to make predictions or classifications in real time. While companies like OpenAI and Meta train massive models using NVIDIA H100s or Google TPUs, the actual deployment of these models, like voice processing, object recognition, or smart replies, can be handled by more efficient hardware.

Apple already uses this kind of hardware for features like:

• Visual Look Up in Photos
• Siri’s on-device responses
• Language translation
• Personalisation in Apple Music and News

This makes Apple Silicon ideal for cloud-based inference workloads, particularly for apps that don’t require the scale or flexibility of AWS’s full GPU stack.

Will Apple Actually Launch This?

That remains uncertain. Michael Abbott, Apple’s former VP of Cloud Engineering and one of the key backers of Project ACDC, left the company in 2023. While discussions reportedly continued into early 2024, The Information notes the current status of the project is unknown.

Still, the internal deployment of Apple Silicon in Apple’s own infrastructure suggests that even if Apple never offers this as a public-facing product, it will continue to leverage this capability to enhance its own services.

Final Thoughts

Project ACDC may never become Apple’s answer to AWS. But it’s clear the company is thinking beyond devices, and towards a future where its silicon not only powers iPhones and Macs, but also the infrastructure running the AI-powered apps of tomorrow.

In a world where cloud costs are rising and AI demands are surging, Apple’s chip efficiency and integrated ecosystem could offer a powerful alternative. Whether it becomes a commercial product or remains a behind-the-scenes advantage, it shows Apple is quietly positioning itself for the next wave of computing.